Unilateral (hemiparetic) cerebral palsy (CP) affects an estimated 229,000 individuals in the U.S., with CP rates increasing in young children. Most traditional forms of therapy have proven ineffective, although they remain in widespread use. Two major research reviews (cf. Huang et al, 2010; Sakzewski et al, 2009) conclude that pediatric constraint-induced movement therapy (CIMT) is one of the most promising new therapies, despite wide variation in and vigorous debate about its key features. Two urgent, unresolved issues concern the effects associated with (a) different dosage levels of pediatric CIMT and (b) different types and duration of constraint. The proposed study is a multi-site, randomized controlled trial (RCT) of 144 children (3 - 6 yrs old) with hemiparetic CP. The multiple baseline 2 X 2 RCT study design will compare the effects of 2 dosage levels (3.0 hrs/day for 21 days versus 1.5 hrs/day for 21 days) under 2 distinct constraint conditions (Full-time cast versus Part-time splint). Children will first be studied in a control condition, receivin their usual therapy over 3 months, and then crossover to be randomly assigned to one of 4 CIMT treatment groups: (1) 3.0 hrs/day dosage, Full-time cast; (2) 3.0 hrs/day, Part-time splint; (3) 1.5 hrs/day, Full-time cast; and (4) 1.5 hrs/day, Part-time splint. For all groups, parents wil receive systematic training to help practice and extend their children's new and improved upper extremity skills. Children will be followed 3 months post-treatment. Children's neuromotor competence will be assessed via a comprehensive battery of standardized and innovative tools, administered 5 times over 8 months by individuals blinded to the subjects' control and treatment conditions. The primary outcome measures are the Assisted Hand Assessment (AHA), the Pediatric Motor Activity Log (PMAL), and the modified QUEST. Independent scoring of videotaped sessions and parent reports will supplement lab assessments, along with data transmitted from accelerometers. Specific Aim 1 focuses on comparing the effects of 2 dosage levels of pediatric CIMT. Specific Aim 2 compares the effects of 2 forms and duration of constraint, selected because of competing hypotheses about their impact on children and cortical re-organization. Specific Aim 3 uses clinical diagnostic and baseline data about subjects' individual differences to explore hypotheses about differential response to CIMT, including their levels of engagement during treatment, the types and magnitude of short- and longer-term changes detected, and whether other changes are documented beyond the realm of neuromotor competence (i.e., spillover effects). The results of this study will advance the development of rigorous, evidence-based standards for the practice of CIMT, with the potential for large-scale implementation to produce major benefits for children with CP.